, esposito - characterization of an effective cleaning procedure for aluminum alloys surface...

8/2/2019 , Esposito - Characterization of an Effective Cleaning Procedure for Aluminum Alloys Surface Enhanced Raman Spe

1/26

LAWRENCE

NAT IONAL

LABORATORY

LIVERMORE

Characterization of an

Effective Cleaning Procedurefor Aluminum Alloys:Surface Enhanced RamanSpectroscopy and Zeta

Potential Anal sis

N. J. Cherepy, T. H. Shen, A. P. Esposito, T. MTillotson

June 9, 2004

UCRL-JRNL-20457

Journal of Colloid and Interface Science


2/26

This document was prepared as an account of work sponsored by an agency of the United States

Government. Neither the United States Government nor the University of California nor any oftheir employees, makes any warranty, express or implied, or assumes any legal liability or

responsibility for the accuracy, completeness, or usefulness of any information, apparatus,

product, or process disclosed, or represents that its use would not infringe privately owned rights.Reference herein to any specific commercial product, process, or service by trade name,

trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement,

recommendation, or favoring by the United States Government or the University of California.

The views and opinions of authors expressed herein do not necessarily state or reflect those ofthe United States Government or the University of California, and shall not be used for

advertising or product endorsement purposes.


3/26

Characterization of an Effective Cleaning Procedure for Aluminum Alloys:

Surface Enhanced Raman Spectroscopy and Zeta Potential Analysis

Nerine J. Cherepy,* Tien H. Shen, Anthony P. Esposito and Thomas M. Tillotson

Chemistry and Materials Science Directorate, Lawrence Livermore National Laboratory,

Livermore, CA 94550

Article to be submitted to the

Journal of Colloid and Interface Science

*Corresponding author: MS L-091, Lawrence Livermore National Laboratory, Livermore, CA

94550, Phone: (925) 424-3492, Fax: (925) 423-8772, [email protected]


4/26

ABSTRACT

We have developed a cleaning procedure for aluminum alloys for effective minimization of

surface-adsorbed sub-micron particles and non-volatile residue. The procedure consists of a

phosphoric acid etch followed by an alkaline detergent wash. To better understand the

mechanism whereby this procedure reduces surface contaminants, we characterized the

aluminum surface as a function of cleaning step using Surface Enhanced Raman Spectroscopy

(SERS). SERS indicates that phosphoric acid etching re-establishes a surface oxide of different

characteristics, including deposition of phosphate and increased hydration, while the subsequent

alkaline detergent wash appears to remove the phosphate and modify the new surface oxide,

possibly leading to a more compact surface oxide. We also studied the zeta potential of


5/26

1

1. INTRODUCTION

Aluminum and stainless alloys have been used extensively in the construction of the

National Ignition Facility (NIF) a 192-beam sports arena-sized laser facility at Lawrence

Livermore National Laboratory. The parts and equipment exposed to the laser beam path must

be kept very clean to protect laser optics from damage and prevent contamination of various

coatings. The cleanliness requirements are similar to those achieved in the semiconductor

industry. The NIF cleanliness criteria for any surfaces that are exposed to laser beam path are as

follows:

1. Non-volatile residues (NVR) at less than Level A/10 (< 0.1 g/cm

2

),

2. Particulate contaminants at less than Level 83 ( i.e. no more than 908 particles, >5 m in

size, per square foot of flat surface) per MIL-STD-1246C specification [1].

To validate these cleanliness criteria, the NVR is measured by a) rinsing the metal surface with

solvent, b) collecting the solvent/dissolved residue, c) evaporating the solvent, and d) weighing

the residue. The particulate contaminant level is measured by the particle swipe test. This test

uses white filter paper to swipe the metal surface and counts the numbers of collected particles

under an optical microscope equipped with a fully automated counting system.

There are many large vacuum facilities requiring high-cleanliness conditions, such as

particle accelerators. Literature concerning the cleaning procedures used for vacuum chamber

construction materials describe the use of alkaline detergent washing followed by deionized

water rinsing for aluminum alloys (see for example [2-5]). However to our knowledge, the use

of a phosphoric acid etch prior to alkaline detergent washing has not been previously described

as a standard cleaning protocol for aluminum vacuum chamber parts, nor has the problem of

smut (loosely attached oxide and surface adsorbed particles) been addressed in this context.


6/26

2

In efforts to clean parts and equipment for the NIF, the parts made out of either cast or

wrought aluminum alloys encountered great difficulty. These early experiences showed that

even with several rigorous high pressure spray washes of alkaline detergent, the aluminum parts

may pass the particle swipe test initially, but fail the test several days later in a Class 100 clean

room. Close examination of the collected particles showed that most of the reappearing particles

are metallic and micron-sized or smaller. Repetitive hand-wipes of the aluminum surface with a

polar solvent, such as isopropyl alcohol, after the detergent wash helped to remove these sub-

micron particles and eventually brought down the particle swipe value to below Level 83.

However, considering the amount of equipment and parts that need to be cleaned, the hand-

wiping after high pressure detergent wash certainly was not an economically acceptable cleaning

procedure.

The chief remedy for particle contamination on aluminum surfaces has been the

implementation of a phosphoric acid etching step prior to the detergent cleaning process [6].

This procedure has been clearly demonstrated on several alloys and parts to effectively minimize

residual particles [7,8]. This multi-step cleaning process consists of a) an initial deionized water

rinse, b) a 30 minute phosphoric acid etch, c) a second deionized water rinse, followed by d) a

5% Brulin 1990GD detergent high pressure spray wash, (or ultrasonic wash with Brulin

815GD for small parts) and e) final rinse with deionized water. The efficacy of the phosphoric

acid etching may arise from: (1) its ability to dissolve small surface-adsorbed particles, (2)

formation of phosphate and/or metaphosphate species at the surface, and (3) surface oxide

modification, by conversion to a different predominant oxide structure and/or change in


7/26

3

morphology (e.g. surface area, compactness, etc.). Either (2) or (3) may modify the surface

electrostatic attraction for residual particles [7].

We have postulated that when the aluminum parts are wet, surface adsorption of sub-

micron particles is enabled via electrostatic attraction to the hydrated aluminum oxide surface

[8]. We have observed that surface-adsorbed metallic particles are readily liberated by solvent

wiping and that the removal of particles may be accelerated by drying with hot air. These two

facts suggest a dehydration mechanism is responsible for the release of the strongly attached

particles.

SEM micrographs of an AA6061-T6 aluminum surface before, versus after, the

phosphoric acid etching are shown in Figure 1. Before the phosphoric acid etching, there are

many sub-micron particles attached to the surface even after several rigorous high pressure

detergent washes. The majority of these particles are aluminum particles; some -eutectic

inclusions (formed by impurities Fe, Si, etc.), iron oxides and stainless steel particles are also

found. The origin of these particles is thought to arise from machining, caustic etching as well as

contamination in the wash water during cleaning. As shown in Figure 1B, many of these

particles disappeared after the phosphoric acid etch and detergent wash; the aluminum surface

was also etched slightly by the acid. However, experimental results showed that the particle

swipe value often increased right after the phosphoric acid etch [6] indicating that more debris

were generated during the etching process. Fortunately, this debris could be easily washed away

by the subsequent alkaline detergent cleaning process, and the reattachment of foreign metallic

particles was prevented. Thus, the phosphoric acid not only dissolved the aluminum particles,

but also in some way changed the nature of the aluminum surface, enabling a more effective

washing step. For aluminum alloys, the effectiveness of the cleaning procedure has been


8/26

4

established in practice, but the mechanism whereby it removes surface-adsorbed particles is the

subject of this work.

The native oxides of aluminum are expected to dominate the surface of most aluminum

alloys, and to control the surface interactions. The passive layer is very thin, consisting of a ~1

nm barrier layer of Al2O3 at the aluminum surface and a porous, typically hydrated

hydroxide/oxyhydroxide layer extending 5-10 nm. Corrosion of aluminum occurs at pH < 4 or

pH > 8.5 according to the Pourbaix diagram [9], however the rate of corrosion can vary greatly

depending on the identity of ionic species controlling pH. Phosphoric acid is second only to

hydrofluoric acid in aggressivity in aluminum etching [10]. It is well- known that acid etching

removes and re-establishes the oxide layer, as well as selectively dissolving Mg2Si precipitates

present in AA6061. Moffitt and co-workers found using the elemental surface spectroscopy x-

ray photoelectron spectroscopy (XPS) that while alkaline cleaning alone of AA2024 did result in

diminished Mg at the surface, a nitric/hydrofluoric acid pickle treatment was more effective [11].

In general, the most effective cleaning procedures for Aluminum alloys are those that result in

the thinnest, densest oxide, with the lowest amount of Mg, as the MgO-Al2O3 oxides tend to be

thicker and more porous [2,11]. A thin oxide layer has less surface area, is less hydrated and is

less able to adsorb particles.

The dehydration of aluminum oxide species has been studied by thermal gravimetric

analysis [12]. Documented dehydration transitions (in air) include: (a) Al(OH)3 AlOOH at

220-230 C, (b) AlOOH - Al2O3 at 310-325 C, (c) AlOOH - Al2O3 at 450-525 C.

Some water seems to be present in samples up to 898 C, finally desorbing completely by 1075

C (shown via IR spectra/dehydration study, Rouquerol, et al. [13]). Under vacuum, the


9/26

5

dehydration of Al(OH)3 starts at 170 C. However, the dehydration of aluminum hydroxide at

room temperature, under vacuum over long periods of time, is not well understood.

In order to examine surface interactions resulting from the cleaning process on the aluminum

surface, we used a variety of surface-sensitive techniques. In earlier work, we employed XPS to

study the aluminum surface. The XPS result [7] suggested the presence of phosphate species on

aluminum surface after a phosphoric acid etch.

We describe here the use of Surface Enhanced Raman Spectroscopy (SERS) to study the

evolution of surface oxide at each cleaning step. SERS is a vibrational spectroscopic technique

that is sensitive only to the surface. Its surface sensitivity derives from a very thin porous

deposit of silver to enhance the Raman spectrum only of the surface it contacts [14]. Raman

spectroscopy provides additional details regarding the molecular structure of the surface (not just

the elemental composition) and it is nondestructive to the oxide structure. SERS allowed us to

characterize the surface species as a function of surface preparation. The methodology for SERS

applied to aluminum with chromium phosphate conversion coatings has been described by

Ahern, et al., at Alcoa [15].

We also describe the use of zeta potential analysis to measure how the surface

electrostatics are modified as a function of surface preparation and wash solution conditions.

Swipe tests show high particle counts if surfaces undergo phosphoric acid etching only, or

detergent wash only, but low residual particles counts when the phosphoric acid etch step

precedes the detergent wash [6]. Zeta potential measurements can indicate whether the etch

changes the surface electrostatics or changes the surface interactions with the detergent.


10/26

6

2. MATERIALS AND METHODS

2.1 Surface Species Identification.

Three coupons of Aluminum Alloy 6061 were prepared for SERS examination as

follows:

1. Sample 1- machined surface wiped with acetone,

2. Sample 2- machined surface etched in phosphoric acid (30 vol% for 30 minutes, then

rinsed in DI water,

3. Sample 3- machined surface etched in phosphoric acid (30 vol %) for 30 minutes,

ultrasonic-cleaned with 3% Brulin 815GD (Brulin Corp., Indianapolis, IN) and 0.02%

Zonyl (DuPont, Wilmington, DE) at 55C for 20 min., followed by DI water rinse.

The sputter deposition of ~5 nm thick Ag coatings for the SERS measurements was

conducted using planar magnetrons operated in the DC mode. When the base pressure of the

vacuum chamber reached a value less than 0.0001 Pa, a sputter gas of Ar was flowed at a

pressure of 0.65 Pa using a flow rate of 40 cc per min. The substrates were positioned 10 cm

above the 6.3 cm diameter sputter target of 0.9999 pure Ag. The magnetron was operated with a

forward power density of 0.15 Watts per cm sq. Each substrate was sequentially exposed to the

shuttered deposition source 20 times over a total time period of 330 sec to yield an average

deposition rate of ~0.015 nm per sec.

Excitation for Raman scattering was provided by the 488 nm line of an argon-ion laser

(Lexel model 95). The laser beam was directed through a narrow bandpass filter to remove the

residual laser plasma lines, and was then coupled into the entrance port of a Raman microscope

(Jobin-Yvon T64000). A beamsplitter partially reflected the laser beam to a 20x objective lens

(Nikon SLWD, NA 0.35) that focused the light to an 8-10 m spot at the sample surface, and the


11/26

7

same objective lens was used to collect and collimate the Raman scattered light obtained at a

180 backscattering geometry. The incident laser power employed was 450 mW as measured at

the sample with the objective lens removed.

The samples were placed on a motorized stage for positioning control, and were visually

inspected using the optical microscope and video monitor, with the laser beam attenuated by a

neutral density filter. Once a sample region was selected for analysis, the neutral density filter

was removed and the scattered light was directed into the triple-grating spectrometer. The

premonochromator was employed to remove the elastic scattering, and final dispersion of the

scattered light was accomplished by a classically ruled 600 grooves/mm (blaze = 500 nm)

grating. The dispersed light was detected with a liquid-nitrogen-cooled CCD camera (Jobin-

Yvon Spectrum-1). Spectra were collected for 30 to 60 seconds, and were averaged ten times for

a total integration time of 5-10 min for each sample region. All spectra were calibrated against a

cyclohexane standard.

2.2 Surface Electrostatics Characterization.

Aluminum particles were obtained from Valimet (Stockton, CA). Two types of particles

were studied, AA6061 (particle size 4-12 m) and pure aluminum (average particle size 2 m).

They were studied as-received and suspended in solution by grinding for 1-2 minutes with

mortar and pestle. To simulate phosphoric acid etching, particles were suspended in 30 vol%

phosphoric acid for 5 minutes, then centrifuged, supernatant discarded, particles rinsed in DI

water. Rinsing in DI water and centrifugation continued until pH registered neutral, usually 3

centrifugation cycles. After the final centrifugation cycle, particles were suspended in a solution

of interest, either 5% Brulin 1990GD in DI water (pH =10.6), or plain DI water with pH


12/26

8

adjusted to 10.6 with sodium disilicate. The detergent composition of Brulin 1990GD is very

similar to that of Brulin 815.

Zeta potential measurements were carried out using a Brookhaven Instruments ZetaPlus

Analyzer. It uses electrophoretic light scattering and laser Doppler velocimetry to determine the

zeta potential of particles suspended in solution. For all reported measurements at least 4

different samples were prepared and each was subjected to 20 light scattering measurements to

generate enough data for statistical analysis.

3. RESULTS AND DISCUSSION

As shown in Figure 1, the phosphoric acid etching modified the AA6061 surface

physically, minimized surface roughness and dissolved small surface-adsorbed particles. To

determine whether phosphoric etching also changes the surface chemically, SERS was

employed. We also probed the surface electrostatics using Zeta Potential Analysis to determine

whether chemical modification of the surface resulted in changes in the strength of potential

interactions between the surface and the particles.

3.1 Surface Species Identification.

SERS spectra were acquired using Raman microscopy, which provides the additional

advantage of an optical image of the oxide deposits overlaid with spectral imaging for

identification of the surface species. A photograph of Sample 3 is shown in Figure 2A, with the

area where silver was deposited apparent as a dark film (porous and ~5 nm thick) over the central

portion of the sample. A magnified optical image of this sample is shown in Figure 2B, showing

machining grooves and pitting due to dissolution of Mg2Si precipitates. Raman spectra were

acquired of both the smooth areas and of the pitted areas. We attempted to obtain spectra from


13/26

9

regions of the aluminum without silver, but within the signal-to-noise of our measurement we

could detect no Raman scattering.

We obtained SERS spectra of each sample in a dark or pitted region, as well in a

smooth region. The spectra for a particular sample were similar, regardless of the surface

characteristics (pitted or smooth), but the spectra of the pitted regions were much more intense,

probably due to a thicker oxide in the pits. Figure 3 shows the spectra of Samples 1, 2 and 3. At

least 4 different spectra for each sample and region type were acquired and found to be

consistent with respect to spectral intensities. The spectra acquired in a pitted region and in a

smooth area of Samples 2 and 3 were subtracted from the spectra acquired for Sample 1, and

shown in Figures 4A and 4B, respectively.

Table 1 lists the vibrational modes of aluminum species that are known from the

literature [15,16]. The SERS spectra of the three AA6061 samples show generally the same

modes, with the exception of two modes at 965 and 1040 cm-1

, present in Samples 2 and 3.

These modes are in good agreement with the frequencies typically exhibited by phosphate

stretching modes as shown in Table 1. Ahern and co-workers assigned modes at 960 and 1055

cm-1

to aluminum phosphate [15]. Other studies point out that the vibrational frequencies of the

isolated [PO4]3-

anion are at ~970 and ~1080 cm-1

, but these modes in AlPO4 appear at higher

frequencies, ~1069-1244 cm-1 [17]. This suggests that the phosphate modes detected via SERS

correspond to very loosely bound, surface adsorbed phosphate. The intensity of these two modes

is much greater for Sample 2, suggesting that the subsequent detergent wash is effective in

removing the phosphate species from the surface.

The 1250-1700 cm-1

region shows more intensity for the phosphoric acid etched sample

than for the other two samples. Primarily this is due to water molecules adsorbed on aluminum,


14/26

10

as assigned by Ahern, et al. This suggests that the phosphoric acid etch promotes hydration of

the aluminum surface to form aluminum hydroxide in addition to excess surface adsorbed water.

Ahern and co-workers also assign some of the intensity in this region to degraded organic

materials (graphitic carbon) [15]. Indeed, some intensity appears to be lost in Sample 3 after the

detergent wash, at ~1333 and 1580 cm-1. Since both water and graphitic carbon have Raman

modes in the same spectral region, this loss of intensity may be due to in part to removal of

organics, but it seems unlikely that without an obvious contamination source that significant

organics would be present on the freshly etched surface. Thus, the reduction in intensity in this

region implies either surface dehydration or the removal of loosely attached hydroxide from the

surface.

Relative to Sample 3, the spectrum of Sample 2 shows more intensity in the modes at

644, 1163, 1290, 1452, 1535 and 1600 cm-1. Meanwhile, Sample 3 shows more intensity in

modes at 772, 815, and 1388 cm-1

. Based on mode assignments as shown in Table 1, these

differences indicate that Sample 2 contains relatively more octahedral aluminum (AlO6) than

tetrahedral aluminum (AlO4) compared to Sample 3. This seems to indicate a different type of

oxide is formed upon phosphoric acid etching, and that it is preferentially removed or

transformed during the detergent wash step.

3.2 Surface Electrostatics Characterization.

We studied AA6061 and pure Al micron-sized particles in solution to try to understand how

particles may adsorb at the surface via local attractive potentials. Surface-adsorbed aluminum

particles, typically arising from machining debris of the bulk material, are negatively charged at

the wash pH (~10.6), as is the bulk surface due to the hydroxide surface structure. These

particles may interact electrostatically with the hydrated alumina surface via adsorbed cations.


15/26

11

The more negative the zeta potential of the particle and the surface, the greater the probability

that strong attractive interactions may occur through the mediation of surface adsorbed cations.

In addition, the more physically and chemically heterogeneous the surface, and the higher the

hydroxide surface area, the greater the likelihood that local attractive potentials may form.

We measured the zeta potential for unetched particles of AA6061 and high purity Al in

DI water (pH adjusted to 10.6 with sodium disilicate) and in 5% Brulin 1990GD. We

performed the same measurements with particles that were surface etched with phosphoric acid.

The results are shown in Table 2 and Figure 5. Unetched AA6061 particles show a decrease of

about 20 mV in the absolute magnitude of the zeta potential in the presence of the detergent,

from -72 to -51 mV. However, high purity Al particles did not exhibit a significant change in

zeta potential in the presence of detergent. Etched particles show a larger in decrease the zeta

potential magnitude with detergent. The magnitude of the zeta potential of both AA6061 and

pure Al particles increases upon etching, and the magnitude of the zeta potential decreases in

Brulin solution, for both types of particles. The magnitude of the zeta potential of etched

AA6061 particles decreases in detergent solution, from -79 mV in pH-adjusted DI water to -42

mV in Brulin solution; a similar decrease is observed for etched pure Al particles, from -70 to -

47 mV.

Our findings indicate that the phosphoric acid etching step results in multiple changes

that may improve the ability of the detergent to adsorb or interact at the surface. They are:

1. A reduction in the amount of surface absorbed metallic particles due to etching and

dissolution of particles,

2. A physically smoother surface at the scale of the machining grooves,

3. Deposition of phosphate species at the surface,


16/26

12

4. Formation of different oxide species at the surface,

5. Increased relative hydration of the oxide at the surface.

Other chemical changes that we did not measure but have been observed in other studies of

aluminum cleaning, such as selective removal of Mg from the surface [4,11] are also expected to

alter the surface area and the surface electrostatics.

Brulin 1990GD, and Brulin 815GD are similar alkaline detergent formulations

recommended for cleaning and degreasing metal parts. They are composed of a proprietary

blend of anionic and non-ionic surfactants, along with some alkaline components and corrosion

inhibitors. Cleaning of aluminum alloy 6xxx series has been found previously to be most

effective with mildalkaline detergents, such as Almeco 18; surfaces cleaned with such

detergents were found to have a thinner oxide, well depleted of Mg, compared to surfaces

cleaned with strong alkaline detergents or solvents only [3]. Moffitt and co-workers found that

an acid pickle was highly effective in removing Mg [11]. The effectiveness of our cleaning

procedure in minimizing surface-adsorbed particles is likely due to (1) formation of a dense, thin

barrier layer at the surface by the phosphoric acid etch, (2) detergent wash removing residual

loosely attached oxide resulting from the acid etch, (3) relative dehydration of the surface during

the detergent wash step, (4) removal of hydrated cations from the surface during the detergent

wash.

4. SUMMARY

An effective cleaning procedure has been developed for aluminum alloys to minimize surface-

adsorbed sub-micron particles and non-volatile residue. This procedure consists of a phosphoric

acid etch followed by a wash with mild alkaline detergent. The SERS measurements indicate


17/26

13

that phosphoric acid etching chemically modifies the surface, depositing phosphate species, and

resulting in an oxide layer with predominant octahedral alumina, as well as an overall higher

degree of hydration. The subsequent alkaline detergent wash then removes phosphate, loosely

attached surface oxide, and dehydrates the surface oxide layer, and leaves behind predominantly

tetrahedral alumina. These changes are indicated by a decrease in the intensity of the phosphate

modes at 965 and 1040 cm-1

and the Al-H2O modes in the 1380-1610 cm-1

range, and a relative

increase in the tetrahedral alumina modes at 772, 815, and 1388 cm -1. The phosphoric acid

etched AA6061-T6 particles exhibited a greater reduction in zeta potential (reduced from -79

mV to -42 mV) in the presence of alkaline detergent, compared to unetched AA6061-T6

particles (-72 mV to -51 mV). These measurements indicate that the newly etched surface

appears to have a chemical and physical morphology that allows it to interact more strongly with

the detergent, resulting in effective removal of surface-adsorbed water, cations, loosely-attached

oxide, and inorganic surface-adsorbed particles.

ACKNOWLEDGEMENTS

We wish to thank William H. Gourdin of the NIF Non-Optical Materials Group for his support

of this work. Thanks also to Alan Jankowski and Jeff Hayes for the silver deposition. We are

grateful to Dan Farber for the use of his Raman instrumentation. The information provided by

Brulin Corp. regarding the formulation of their products was of great help. This work was

performed under the auspices of the U.S. Department of Energy by the University of California,

Lawrence Livermore National Laboratory under contract No. W-7405-Eng-48.


18/26

14

REFERENCES

(1) MIL-STD-1246C, Product Cleanliness Levels and Contamination Control Program

Institute of Environmental Sciences and Technology, 940 E. Northwest Highway, Mt.

Prospect, IL 60056.(2) Y. Tito Sasaki, J. Vac. Sci. Technol. A, 9 (1991) 2025.

(3) R.A. Rosenberg, M.W. McDowell, J.R. Noonan, J. Vac. Sci. Technol. A, 12 (1994) 1755.(4) C.L. Foerster, C. Lanni, R. Lee, G. Mitchell, W. Quade, J. Vac. Sci. Technol. A, 15

(1997) 731.

(5) C. Benvenuti, G. Canil, P. Collin, R. Cosso, J. Guirin, S. Ilie, D. Latorre, K.S. Neil,Vacuum, 53 (1999) 317.

(6) T. Shen and M. Fischer, The Cleaning of Wrought Aluminum Alloy 6061-T6, LLNL

report dated June 29, 2001, UCRL-ID-145445.

(7) T. Shen, The Cleaning of OAB Universal Covers- an Origin of Smut in AluminumAlloys, LLNL report dated June 2, 2002, UCRL-ID-148327.

(8) T. Shen, The Gross Cleaning of Aluminum Alloys The Effect of Caustic Etch on SmutFormation, LLNL report dated August 30, 2002, UCRL-MI-149969.(9) D.A. Jones, Principles and Prevention of Corrosion, Macmillan Publishing Co, New

York, 1995.

(10) J.R. Davis, Ed., Corrosion of Aluminum and Aluminum Alloys, ASM Intl., 1999.(11) C.E. Moffitt, D.M. Wieliczka, H.K. Yasuda, Surf. Coat. Technol., 137 (2001) 188.

(12) J.W. Diggle, A.K. Vijh, Eds., Oxides and Oxide Films, Marcel Dekker, Inc., 1976.

(13) J. Rouquerol, J. Fraissard, J. Elston, B. Imelik, J. Chim. Phys., 63 (1966) 607.

(14) K. Kneipp, H. Kneipp, I. Itzkan, R.R. Dasari, M. S. Field, Chem. Rev., 99 (1999) 2957.(15) A.M. Ahern, P.R. Schwartz, L. A. Shaffer, Appl. Spectr., 46 (1992) 1412.

(16) T. Schram, J. DeLaet, H. Terryn, J. Electrochem. Soc., 145 (1998) 2733.

(17) M. Rokita, M. Handke, W. Mozgawa, J. Mol. Struct., 555 (2000) 351.


19/26

15

TABLES

Table 1. Vibrational mode assignments of observed Raman transitions.

Present work

(cm-1

)

Previously assigned

modes (cm-1

)

Bond

617/644 626/630a

AlO6 (octahedral)

772 668/772/731a

AlO4 (tetrahedral)

815 834 a, 840 b AlO4

928/934 a, 916 b Al-O- Al asymmetric stretch

965 960 a Phosphate stretch

1056/1068 a Al- O bending mode

1040 1055/1117a

Phosphate

1163

1290 1290b

Aluminum oxide mode

1382/1358a

carbon

1584/1582a

carbon

1388 1360a

H2O coordinated to AlO4

1452 1425a

H2O coordinated to AlO6

1535

1600 1611 a H2O coordinated to AlO6

1645 a physisorbed H2O

afrom ref. 14bfrom ref. 15


20/26

16

Table 2. Zeta potentials measured for particles in either deionized water or detergent.

Particle Type Pretreatment Solution (mV)

AA6061 unetched DIa -71.86 4.25

High purity Al unetched DIa

-61.50 7.97

AA6061 unetched detergentb

-51.43 4.30

High purity Al unetched detergentb

-60.74 5.10

AA6061 H3PO4 etched DIa

-79.26 1.65

High purity Al H3PO4 etched DIa

-70.17 7.60

AA6061 H3PO4 etched detergentb

-41.99 7.10

High purity Al H3PO4 etched detergentb

-46.98 4.96

apH adjusted to 10.6 with sodium disilicate

b5% Brulin 1990GD, pH= 10.6


21/26

17

FIGURE CAPTIONS

Figure 1. (A) SEM micrograph of an AA6061-T6 aluminum surface. (B) Same surface, after

the phosphoric acid etch and detergent cleaning, exhibits smoother features, and less adsorbed

particles.

Figure 2. (A) Photograph of aluminum coupon coated with silver in the middle portion for SERS

measurements. (B)Close -up image of the surface of Sample 1, shows pitted and smooth surface

areas, as well as grooves from machining.

Figure 3. SERS spectra acquired in pitted regions at the surfaces of Samples 1, 2 and 3.

Figure 4. (A) SERS difference spectra showing the spectrum acquired for Sample 1 subtracted

from the spectra of Samples 2 and 3, all data acquired in pitted regions (difference spectra

generated from data shown in Figure 3). (B) SERS difference spectra acquired in smooth surface

regions.

Figure 5. Averaged zeta potential measurements of pure aluminum and AA6061 particles with

error bars.


22/26

18

FIGURES

Figure 1A, 1B

30 m_____

B


23/26

19

Figure 2A

Figure 2B

Dark- pit/crevice, thick oxide

Smooth- thin oxide

10 m


24/26

20

Figure 3

25x103

20

15

10

I

1600140012001000800600

Raman shift (cm-1

)

644

1290

1388

1452

1600

772

815

SERS, pitted region

1163

1040

1535

965

617

Sample 1 (acetone wipe)

Sample 2 (phosphoric etch)

Sample 3 (etch + Brulin wash)


25/26

21

Figures 4A, 4B

5000

4500

4000

3500

3000

2500

2000

1500

1000

500

0

I

644

12901388

1452

1600

772

815

SERS, pitted region

1163

1040

1535

965



600

400

200

0

I

1600140012001000800600

Raman shift (cm-1

)

SERS, smooth region

965 1040

1388 1535

1600




26/26

Figure 5

-80

-70

-60

-50

-40

-30

ZetaP

otential(mV)

6061,

DI

Al,DI

6061,

Brulin

Al,Brulin

unetched

etched

, esposito - characterization of an effective cleaning procedure for aluminum alloys surface...

Documents